Wide dynamic range active pixel with knee response

Abstract

A pixel circuit, and a method for operating a pixel circuit, to provide a multiple knee response characteristic. In one embodiment a pixel circuit comprises a photoconversion device for accumulating charge during a first integration period and second integration period, an integration node connected to the photoconversion device, a first transistor having one terminal connected to said integration node and another terminal connected to a reset signal line and a feed-through pulse capacitor. The feed-through pulse capacitor has one terminal coupled to a feed-through pulse signal line, and a second terminal coupled to the integration node, said feed-through pulse signal line providing an intermediate pulse between the first and second integration periods to generate an overflow current in said the transistor.

Description

CROSS REFERENCE TO RELATED APPLICATIONS

This application is a continuation of U.S. patent application Ser. No. 10/431,397, filed on May 8, 2003 now U.S. Pat. No. 7,224,388, the entire disclosure of which is incorporated by reference.

The present invention relates to a method of operating a pixel circuit to increase intrascene dynamic range while reducing fixed pattern noise.

BACKGROUND OF THE INVENTION

Intrascene dynamic range refers to the range of incident light that can be accommodated by an image sensor in a single frame of pixel data. Examples of scenes that generate high dynamic range incident signals include an indoor room with outdoor window, an outdoor scene with mixed shadows and bright sunshine, night-time scenes combining artificial lighting and shadows and, in an automotive context, an auto entering or about to leave a tunnel or shadowed area on a bright day.

Dynamic range is measured as the ratio of the maximum signal that can be meaningfully imaged by a pixel of the imager to its noise level in the absence of light. Typical CMOS active pixel sensors (and charge coupled device (CCD)) sensors have a dynamic range from 60 to 75 dB. This corresponds to light intensity ratios of 1000:1 to about 5000:1. Noise in image sensors, including CMOS active pixel image sensors, is typically between 10 and 50 e-rms. The maximum signal accommodated is approximately 30,000 to 60,000 electrons. The maximum signal is often determined by the charge-handling capacity of the pixel or readout signal chain. Smaller pixels typically have smaller charge handling capacity.

Typical scenes imaged by cameras have lighting levels that generate signals on the order of 10-1,000 electrons under low light (1-100 lux), 1000-10,000 electrons under indoor light conditions (100-1000 lux), and 10,000->1,000,000 electrons (1000-100,000 lux) under outdoor conditions. To accommodate lighting changes from scene to scene, the so-called intrascene dynamic range, an electronic shutter is used to change the integration time of all pixels in the arrays from frame to frame.

To cover a single scene that might involve indoor lighting (100 lux) and outdoor lighting (50,000 lux), the required intrascene dynamic range is of the order of 5,000:1 (assuming 10 lux of equivalent noise) corresponding to 74 dB. In digital bits, this requires 13-14 bits of resolution. However, most CMOS active pixel sensors have only 10 bits of output and 8 bits of resolution typically delivered to the user in most image formats such as JPEG. Companding of the data is often used to go from 10-12 bits to 8 bits. One type of companding is gamma correction, where roughly the square root of the signal is generated.

In order to accommodate high intrascene dynamic range, several different approaches have been proposed in the past. A common denominator of most approaches is performing signal companding within the pixel by having either a total conversion to a log scale (so-called logarithmic pixel) or a mixed linear and logarithmic response in the pixel. One example of a mixed linear and logarithmic circuit can be found in co-pending and commonly assigned patent application Ser. No. 10/226,127, filed Aug. 23, 2002, titled “A Wide Dynamic Range Linear-And-Log Active Pixel,” the disclosure of which is incorporated by reference herein.

These prior approaches have several major drawbacks. First, the knee point in a linear-to-log transition is difficult to control, leading to fixed pattern noise in the output image. Second, under low light, the log portion of the circuit is slow to respond, leading to lag. Third, a logarithmic representation of the signal in the voltage domain (or charge domain) means that small variations in signal due to fixed pattern noise leads to large variations in the represented signal.

Linear approaches have also been described where the integration time is varied during a frame to generate several different signals. This approach has architectural problems if the pixel is read out at different points in time since data must be stored in an on-board memory before the signals can be fused together. Another approach is to integrate two different signals in the pixel, one with low gain and one with high gain. However, the low gain portion of the pixel often presents color separation processing problems.

BRIEF SUMMARY OF THE INVENTION

The present invention relates to increasing intra-scene dynamic range for image capturing in a pixel circuit. In one aspect, the invention provides a pixel circuit having an integration node; a conversion transistor having a source/drain connected to the integration node and a drain/source connected to a reset line; a feed-through pulse capacitor having one leg connected to a feed-through pulse signal line and the other leg connected to the integration node; a photodiode having one leg connected to the integration node; and an output transistor having a gate connected to the integration node.

The pixel circuit of the present invention can be operated such that one or more feed-through pulse (FTP) signals are transmitted to an integration node after a first linear integration time period. After the FTP signal is transmitted, a second linear integration period is initiated on the pixel circuit, where the photo conversion gain of the pixel circuit becomes reduced due to leaked electrons from the photodiode. By generating overflow current in a transfer transistor during medium and high illumination conditions in one embodiment of the invention, the pixel circuit can operate with less noise. Such operation creates a pixel with a photo response having multiple “knee” points, where each “knee” in the photo response curve creates separate regions with photo sensitivities that can be independently controlled with minimal thermal interference.

The disclosed configuration further provides added flexibility to controlling the photo response of a pixel circuit. By adjusting the FTP signal voltage (e.g., high, medium, low) and/or the integration time period, the photo response may be more easily suited to the needs of users.

In another aspect, the invention provides a method of operating the pixel circuit, where an overflow pulse is transmitted on a feed-through pulse (FTP) line during a blanking period. During the blanking period, scanned data is not transmitted; thus noise from the FTP line is significantly reduced in the resulting video signal.

In yet another embodiment, a FTP signal is used in conjunction with a readout signal RD in the pixel circuit to further stabilize knee responses during a multiple-knee operation during an overflow condition.

These and other features and advantages of the invention will be more clearly seen from the following detailed description of the invention which is provided in connection with the accompanying drawings.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a block diagram of an imaging device of an embodiment of the present invention;

FIG. 1A is a schematic diagram of an active pixel circuit of an embodiment of the present invention;

FIG. 2 is an exemplary timing diagram illustrating the operation of the FIG. 1A pixel circuit under low illumination;

FIG. 3 partially illustrates the pixel circuit of FIG. 1A fabricated on a semiconductor substrate;

FIG. 3A is a potential distribution diagram relating to the circuit of FIG. 3, illustrating a signal level readout condition on the reset region and the photodiode region under a low illumination condition;

FIG. 3B is a potential distribution diagram relating to the circuit of FIG. 3, illustrating a bias charge injection condition on the reset region and the photodiode region under a low illumination condition;

FIG. 3C is a potential distribution diagram relating to the circuit of FIG. 3, illustrating a bias charge overflow condition on the reset region and the photodiode region under a low illumination condition;

FIG. 3D is a potential distribution diagram relating to the circuit of FIG. 3, illustrating a reset level readout condition and the beginning of a first integration on the reset region and the photodiode region under a low illumination condition;

FIG. 3E is a potential distribution diagram relating to the circuit of FIG. 3, illustrating an overflow condition on the reset region and the photodiode region under a low illumination condition;

FIG. 3F is a potential distribution diagram relating to the circuit of FIG. 3, illustrating a signal level readout condition on the reset region and the photodiode region under a low illumination condition;

FIG. 3G is a potential distribution diagram relating to the circuit of FIG. 3, illustrating the completion of a second integration on the reset region and the photodiode region under a low illumination condition;

FIG. 4 is an exemplary timing diagram illustrating an operating condition of the FIG. 1A pixel circuit under medium illumination;

FIG. 5 partially illustrates the pixel circuit of FIG. 1A fabricated on a semiconductor substrate;

FIG. 5A is a potential distribution diagram relating to the circuit of FIG. 3, illustrating a signal level readout condition on the reset region and the photodiode region under a medium illumination condition;

FIG. 5B is a potential distribution diagram relating to the circuit of FIG. 3, illustrating a bias charge injection condition on the reset region and the photodiode region under a medium illumination condition;

FIG. 5C is a potential distribution diagram relating to the circuit of FIG. 3, illustrating a bias charge overflow condition on the reset region and the photodiode region under a medium illumination condition;

FIG. 5D is a potential distribution diagram relating to the circuit of FIG. 3, illustrating a reset level readout condition and the beginning of a first integration on the reset region and the photodiode region under a medium illumination condition;

FIG. 5E is a potential distribution diagram relating to the circuit of FIG. 3, illustrating an overflow condition on the reset region and the photodiode region under a medium illumination condition;

FIG. 5F is a potential distribution diagram relating to the circuit of FIG. 3, illustrating a signal level readout condition on the reset region and the photodiode region under a medium illumination condition;

FIG. 5G is a potential distribution diagram relating to the circuit of FIG. 3, illustrating the completion of a second integration on the reset region and the photodiode region under a medium illumination condition;

FIG. 6 is an exemplary timing diagram illustrating an operating condition of the FIG. 1A pixel circuit under high illumination;

FIG. 7 is a graph illustrating the photo response of the pixel circuit of FIG. 1A under a high illumination condition;

FIG. 8 is an exemplary timing diagram illustrating an operating condition of the FIG. 1A circuit using a multiple-knee response under an alternate embodiment of the invention;

FIG. 9 is a graph of the photo response of the FIG. 1A pixel circuit utilizing the timing diagram of FIG. 8;

FIG. 10 is a block diagram of an exemplary imager utilizing the knee response pixel of FIG. 8 under another embodiment of the invention;

FIG. 11 is an exemplary timing diagram of the imager of FIG. 10; and

FIG. 12 depicts a block diagram of a processor system employing the FIG. 1A pixel circuit, in accordance with yet another exemplary embodiment of the invention.

DETAILED DESCRIPTION OF THE INVENTION

Embodiments of the present invention are employed in a CMOS imaging device generally illustrated in FIG. 1 by reference numeral 10. The imaging device includes an array of pixels arranged in rows and columns with each pixel having a pixel circuit 100, each pixel being associated with a column line to which all pixels of that column are connected, the pixels being selected row-by-row. The pixel circuit 100 provides a reset signal V_RST and a pixel image signal V_SIG as outputs during a reset and integration period, respectively, which are captured by a sample and hold circuit 200 associated with that column in response to sampling signals SHS (for the image signal) and SHR (for the reset signal), respectively. The sample and hold circuit 200 passes the reset signal V_RST and image signal V_SIG of a pixel circuit 100 to an amplifier 40 which in turn provides a signal representing the difference between the reset signal and pixel image signal (V_RST−V_SIG) as an output. This difference signal is provided to an analog to digital converter 60 and, from there, to an image processor 80 which receives digitized pixel signals from all pixel circuits 100 of the pixel array and provides an image output.

An active pixel circuit 100 in accordance with an embodiment of the invention is shown in greater detail in FIG. 1A. Pixel circuit 100 includes a transfer transistor 116, an output transistor 120, a row select transistor 124, a photodiode 108, and a feed-through capacitor 117. Also provided are a row select signal line 131 receiving a row select signal RD, a reset signal line 121 receiving a reset signal RST and a feed through pulse line 119 receiving a feed through pulse signal FTP. A voltage supply line 123 is also provided which supplies a voltage VAAPIX to the pixel circuit 100.

The transfer transistor 116 has a gate threshold voltage of V_t and is operated either in a shut-off voltage operating mode or a sub-threshold voltage operating mode, as described in greater detail below. The feed through capacitor 117 is located between the horizontal feed through pulse (FTP) signal line 119 and a signal integration node 104. One source/drain region of the transistor 116 is connected to the row reset (RST) signal line 121, while the gate of transistor 116 is connected to the power supply line VAAPIX 123, and the other source/drain region of transistor 116 is connected to integration node VPIX 104. The photodiode 108 is connected to the integration node 104 and ground. One source/drain region of an output transistor 120 is connected to the supply line VAAPIX 123 while the gate of transistor 120 is connected to the integration node 104. The gate of row select transistor 124 is connected to the row select signal line which receives the row select signal RD, while the source/drain regions of the transistor 124 are respectively coupled to output transistor 120 and column line 126.

When connected to the column line 126 through the row select transistor 124 as described above, the output transistor 120 operates as a source follower transistor and provides a gain to the charge signal received from node 104.

As noted, transistor 116 has two operating modes. One operating mode is a shut-off operating mode in which the transistor 116 imparts a linear output to an accumulated pixel image signal V_SIG at node 104 during a charge integration period, while the other operating mode is a sub-threshold operating mode which imparts a logarithmic output to the pixel image signal V_SIG accumulated at node 104. The SHS and SHR pulses correspond to when the signal and reset voltages, respectively, are sampled. As with the control lines FTP, RD, and RST discussed below, the SHR and SHS pulses are produced by the signal controller 70.

The operation of the pixel circuit 100 will now be explained with reference to the timing diagram of FIG. 2, which shows a typical frame cycle during operation of the pixel circuit 100 under a low illumination condition. At time t₁, the sample and hold signal SHS pulse (not shown) initiates a pixel image sampling signal to be applied to a sample and hold circuit which causes the pixel image signal V_SIG to be sampled and held. The read out signal RD at time t₁ is also high, signifying that charge accumulated at a node 104 is being read out. This charge is accumulated at node 104 prior to the time t₁. At time t₂, the RST line and the feed-through pulse line (FTP) go low (shown as V_FTP—L for the feed-through pulse). This causes V_PIX, the voltage at node 104, to be set to the RST line 121 low voltage. At time t₃, the RST line 121 goes high, which begins the process of resetting the pixel. This causes V_PIX, the voltage at node 104, to begin increasing. The voltage V_PIX may be expressed in terms of equation (1), shown below:

$\begin{array}{cc}{V}_{\mathrm{PIX}}\left(t\right)=\frac{1}{\beta }\ln \left[kt+\exp \left(\beta \times V_{\mathrm{PIX}}\left(t_3\right)\right)\right]\mathrm{where}k=\frac{\beta I_0}{C_{\mathrm{pix}}}\exp \left[\beta \left(\mathrm{VAAPIX}-{\mathrm{VT}}_0\right)\right]& \left(1\right)\end{array}$
where β represents an exponential coefficient of the subthreshold current of transistor 116, I₀ represents the subthreshold current of transistor 116, VT₀ represents the subthreshold voltage, and C_pix represents the total capacitance at the node 104. After the resetting operation is initiated, charge from the reset node is subtracted from any prior signal levels, thus significantly reducing or even eliminating offset variation in the pixel.

At time t₄, FTP pulse goes high, causing VPIX to reach the level shown in equation (2) below:

$\begin{array}{cc}{V}_{\mathrm{PIX}}\left(t_4\right)=\frac{1}{\beta }\ln \left[{\mathrm{kt}}_{\mathrm{RST}}\right]+\frac{C_{\mathrm{FTP}}}{C_{\mathrm{pix}}}\left(V_{\mathrm{FTP}\_H}-V_{\mathrm{FTP}\_L}\right)& \left(2\right)\end{array}$

where t_RST represents the overflow reset time (t₄-t₃). The second term of equation (2) represents feed-through charge injected by the FTP pulse, where C_FTP represents the capacitance of capacitor 117, C_pix represents the total capacitance at the node 104, and V_FTP—H and V_FTP—L are the high and low levels of the FTP pulse illustrated in FIG. 2. It should be noted that C_pix consists of C_FTP and also includes the capacitance of the photodiode 108 and the sum of parasitic capacitances of the circuit 100 such as the gate capacitance of the transistor 120, and the junction capacitance of the source node of the transistor 116. Prior to a reset operation, a substantial amount of charge is injected into the pixel capacitor C_FTP and its potential is then pinned at the ‘low’ level of the RST line as shown between the time period t₂-t₃ of FIG. 2. Because of this pimning action, the primary integrated signal is fully discharged from C_pix, so that the reset operation completely resets the circuit 100, and excess charge from previous imaging cycles of the circuit 100 does not ‘lag’ into following imaging cycles.

At time t₅, the RD line goes low, ending the first readout process, thus beginning a charge accumulation (integration) period. During the period from t₅ to t₆, the transfer transistor 116 operates in a shut-off mode and a linear accumulated charge signal is processed at the node 104.

At the point t₆, the FTP signal drops to a medium level (V_FTP—M), which interrupts the integration period. A signal charge, represented by I_ph×tac1, is accumulated at the pixel node V_PIX 104, where I_ph represents the photocurrent present at the node, and tac1 represents the first integration period (t₆-t₅). When the FTP pulse drops to V_FTP—M at t₆, V_PIX reaches the level shown in equation (3) below:

$\begin{array}{cc}{V}_{\mathrm{PIX}}\left(t_6\right)=V_{\mathrm{PIX}}\left(t_4\right)-\frac{I_{ph}\times \mathrm{tac}1}{C_{\mathrm{pix}}}-\frac{C_{\mathrm{FTP}}}{C_{\mathrm{pix}}}\left(V_{\mathrm{FTP}\_H}-V_{\mathrm{FTP}\_M}\right)& \left(3\right)\end{array}$

If the accumulated charge at the pixel node 104 is sufficiently small, the bias transistor MCM 116 will not turn on, and the accumulated charge will remain at node 104. Thus the FTP pulse does not influence the signal charge, and V_PIX returns back to its initial voltage level at t₇, when the FTP pulse goes back to V_FTP—H, as shown in FIG. 2. Once the FTP pulse reaches V_FTP—H at t₇, the integration period resumes, where the voltage V_PIX becomes:

$\begin{array}{cc}{V}_{\mathrm{PIX}}\left(t_7\right)=V_{\mathrm{PIX}}\left(t_4\right)-\frac{I_{ph}\times \mathrm{tac}1}{C_{\mathrm{pix}}}& \left(4\right)\end{array}$

At the time t₁′ the charge accumulation (integration) period ends and the accumulated pixel voltage V_PIX is read out by the transistors 120, 124 as the pixel image signal V_SIG, and a new frame cycle begins. After the end of the second integration period, a charge of I_ph×tac2 is additionally integrated, where I_ph represents the photocurrent at node 104, and tac2 represents the second integration time period signified by t₁′-t₇. The voltage V_PIX at time t₁′ may be expressed as:

$\begin{array}{cc}{V}_{\mathrm{PIX}}\left(t_1^{\prime }\right)=V_{\mathrm{PIX}}\left(t_4\right)-\frac{I_{ph}\times \mathrm{tac}1}{C_{\mathrm{pix}}}-\frac{I_{ph}\times \mathrm{tac}2}{C_{\mathrm{pix}}}& \left(5\right)\end{array}$

The photo response of pixel circuit 100 can thus be expressed as:

$\begin{array}{cc}\begin{array}{c}\mathrm{Sig}=G_{\mathrm{SF}}\times \left(\frac{I_{ph}\times \mathrm{tac}1}{C_{\mathrm{pix}}}+\frac{I_{ph}\times \mathrm{tac}2}{C_{\mathrm{pix}}}\right)\\ =G_{\mathrm{SF}}\times \frac{I_{ph}}{C_{\mathrm{pix}}}\times \left(t_1^{\prime }-t_5\right)\end{array}& \left(6\right)\end{array}$
where G_SF represents the source-follower gain.

Turning to FIG. 3, a portion of the circuit of FIG. 1A is illustrated as an embodiment fabricated on a semiconductor substrate, for example, a silicon substrate. Reset line 121 is shown being connected to reset electrode region 302, which is adjacent to transfer transistor MCM 116. Transistor MCM 116 is further coupled to the VAAPIX line 123 as shown in FIG. 3. The FTP 119 line is connected to capacitor CFTP 100, which connects further to the photodiode region 303, and to the gate of readout transistor 120. One source/drain terminal of readout transistor 120 is coupled to the VAAPIX line, while the other source/drain terminal of transistor 120 is connected to a source/drain terminal of transistor 124. The gate of transistor 131 is connected to the row select line 131, and the other source/drain terminal is connected to the output PIXOUT, and to an external load 301, which has been illustrated as a current source in FIG. 3.

FIGS. 3A-3G illustrate an exemplary potential distribution diagram for the circuit of FIG. 3 under a low-illumination condition, where the potential charge, or electrons 310 between regions 302 and 303 are illustrated. The barrier between the reset region 302 and the photodiode region 303 electrically isolates the photodiode region from the transfer transistor 116 during the integration period. Thus, any photo-generated charge 311 produced by photodiode 303 is initially stored in the right well region. The left-well region associated with reset node 302 is directly connected to reset line RST 121, and stores the charge received from the reset RST 121 line.

Turning to FIG. 3A, the exemplary potential distribution diagram illustrates the potential charge 310 present during the low illumination signal level readout at time t₁, wherein the photodiode region accumulates photo-generated charge 311 after a previous integration period. Since the sum of the potential charge 310 and photo-generated charge potential 311 does not exceed the barrier potential, the potential is held in the diode region 303. At time t₂, both the RST pulse and the FTP pulse go low (see FIG. 2), at which time all regions are filled with electrons 310, via a bias charge, as shown in FIG. 3B.

Turning to FIG. 3C, when the RST pulse goes high at time t₃ (see FIG. 2), a bias charge overflow occurs in the reset region 302, and the excess bias charge is swept away from the reset region 302. If the reset region 302 potential exceeds the barrier potential, the photodiode region 303 potential is pinned at the potential of the reset region by an electrical channel (not shown) formed at the barrier region. Since transistor 116 is operating in the subthreshold region, the overflow current (I_MCM) can be expressed as:
I_MCM=I₀exp[β×(VAAPIX−V_PIX−VT₀)]  (7)
where I₀ represents the sub-threshold current of transfer transistor 116, and VT₀ represents the sub-threshold voltage across transfer transistor 116.

At time t₄, illustrated in FIG. 3D, the feed-through pulse FTP goes high (see FIG. 2), and the reset level is read out and subtracted from the prior readout signal level so that offset variation of the pixel can be eliminated. After the first integration period, when t=t₆ (see FIG. 2), additional photo-generated charge 311 is accumulated at the photodiode region as shown in FIG. 3E. However, since the potential is not great enough under low illumination to overcome the barrier, the charge is held in the photodiode region 303. In FIG. 3F, the additional charge 311 is integrated after t₇ (See FIG. 2), wherein the integration period ends at t=t₁′ (FIG. 3G), and a voltage readout occurs where the voltage from the photodiode region is read by source followed transistor 120 and row select transistor 124 onto the PIXOUT line 126.

Turning to FIG. 4, the exemplary timing diagram shows a typical frame cycle during operation of the pixel circuit 100 under a medium illumination condition. For times t₁ to t₅, the timing operation is substantially identical to the corresponding times illustrated in the low illumination timing diagram of FIG. 2. After time t₅, pixel circuit undergoes an integration period (tac1) under medium illumination. During the integration period under a medium illumination condition, the voltage V_PIX can be expressed as:

$\begin{array}{cc}{V}_{\mathrm{PIX}}\left(t_4\right)-\frac{I_{ph}\times \mathrm{tac}1}{C_{\mathrm{pix}}}& \left(8\right)\end{array}$
where I_ph is the photodiode current, and C_pix is the total capacitance at the integration node 104. When the FTP pulse transitions from V_FTP—H to V_FTP—M at time t₆, V_PIX drops to a lower level (represented by equation (3) above), causing transfer transistor MCM 116 to turn on. Once transistor 116 turns on, the accumulated photo-charge is drained through transistor 116, and the VPIX voltage at time t₇ is:

$\begin{array}{cc}{V}_{\mathrm{PIX}}\left(t_7\right)=\frac{1}{\beta }\ln \left[k\left(t_7-t_6\right)\right]+\frac{C_{\mathrm{FTP}}}{C_{\mathrm{pix}}}\left(V_{\mathrm{FTP}\_H}-V_{\mathrm{FTP}\_M}\right)& \left(9\right)\end{array}$
and if the time period between t₇ and t₆ is set at the same length of the reset time t_RST (t₄-t₃), equation (9) becomes:

$\begin{array}{cc}{V}_{\mathrm{PIX}}\left(t_7\right)=V_{\mathrm{PIX}}\left(t_4\right)-\frac{C_{\mathrm{FTP}}}{C_{\mathrm{pix}}}\left(V_{\mathrm{FTP}\_M}-V_{\mathrm{FTP}\_L}\right)& \left(10\right)\end{array}$

The difference between the circuit 100 operation under low illumination operation versus medium illumination operation can be defined by equation (11) shown below:

$\begin{array}{cc}\frac{I_{ph}\times \mathrm{tac}1}{C_{\mathrm{pic}}}=\frac{C_{\mathrm{FTP}}}{C_{\mathrm{pix}}}\left(V_{\mathrm{FTP}\_M}-V_{\mathrm{FTP}\_L}\right)& \left(11\right)\end{array}$
where, if photodiode current I_ph is larger than the photodiode transition point, excess charge overflows through transistor MCM 166 and a medium illumination condition begins. The photodiode transition point can be expressed as:

$\begin{array}{cc}{I}_{ph\left(\mathrm{transition}\right)}=\frac{C_{\mathrm{FTP}}}{\mathrm{tac}1}\left(V_{\mathrm{FTP}\_M}-V_{\mathrm{FTP}\_L}\right)& \left(12\right)\end{array}$

After t₇, FTP pulse goes back high, and pixel circuit 100 resumes charge accumulation under a second accumulation period (t₁′-t₇=tac2). If the second accumulation period (tac2) is shorter than the first accumulation period (tac1), signal I_ph×tac2 is added to the integration node 104. When the reset pulse RD ends the accumulation period at t1′, VPIX may be expressed as:

$\begin{array}{cc}{V}_{\mathrm{PIX}}\left(t_1^{\prime }\right)=V_{\mathrm{PIX}}\left(t_4\right)-\frac{C_{\mathrm{FTP}}}{C_{\mathrm{pix}}}\left(V_{\mathrm{FTP}\_M}-V_{\mathrm{FTP}\_L}\right)-\frac{I_{ph}\times \mathrm{tac}2}{C_{\mathrm{pix}}}& \left(13\right)\end{array}$

By subtracting the offset from the signal, the photo response of pixel circuit 100 can be expressed as:

$\begin{array}{cc}\mathrm{Sig}=\frac{G_{\mathrm{SF}}}{C_{\mathrm{pix}}}\times \left[C_{\mathrm{FTP}}\times \left(V_{\mathrm{FTP}\_M}-V_{\mathrm{FTP}\_L}\right)+I_{ph}\times \mathrm{tac}2\right]& \left(14\right)\end{array}$

Turning to FIG. 5, a portion of the circuit of FIG. 1A is illustrated as an embodiment fabricated on a semiconductor substrate. The circuit of FIG. 5 is substantially identical to the circuit of FIG. 3, which was discussed above.

FIGS. 5A-G illustrate exemplary potential distribution diagrams for the circuit of FIG. 5 under a medium-illumination condition, where the potential charge, or electrons 310 between regions 302 and 303 are illustrated. Turning to FIG. 5A, the exemplary potential distribution diagram illustrates the potential charge 310 present during the medium illumination signal level readout at time t₁, wherein the photodiode region accumulates photo-generated charge 311 after a previous integration period. At time t₂, both the RST pulse and the FTP pulse go low (see FIG. 4), at which time all regions are filled with electrons 310, via a bias charge as shown in FIG. 5B.

Turning to FIG. 5C, when the RST pulse goes high at time t₃ (see FIG. 4), a bias charge overflow occurs in the reset region 302, and the excess bias charge is swept away from the reset region 302. Transistor 116 is operating in the subthreshold region, thus producing the overflow current (I_MCM) expressed as equation (7), discussed above.

At time t₄, illustrated in FIG. 5D, the feed-through pulse FTP goes high (see FIG. 4), and the reset level is read out and subtracted from the prior readout signal level so that offset variation of the pixel can be eliminated. After the first integration period, when t=t₆ (see FIG. 4), bias transistor MCM 116 turns on, allowing excess photo-generated charge 311 accumulated at the photodiode region 303 to drain through transistor MCM 116 to the reset region 302 as shown in FIG. 5E.

Turning to FIG. 5F, the second integration period is illustrated, where additional photo-generated charge 311 is accumulated at the photodiode region 303. In FIG. 5F, the additional charge 311 is integrated after t₇ (See FIG. 2), wherein the integration period ends at t=t₁′ (FIG. 5G), and a voltage readout occurs where the voltage from the photodiode region is read by source follower transistor 120 and row select transistor 124 onto the PIXOUT line 126.

FIG. 6 shows an exemplary timing diagram of a typical frame cycle during operation of the pixel circuit 100 under a high illumination condition. The timing operation of the FIG. 6 embodiment is substantially identical to the corresponding times illustrated in the medium illumination timing diagram of FIG. 4, except that V_PIX reaches a saturation (overflow) point 601 during the first integration period (tac1), as well as during the second integration period (tac2), illustrated by the dotted line 600 in FIG. 6.

The photo response of pixel circuit 100 is further illustrated in FIG. 7, wherein the photo conversion characteristic has knee points (700, 701) between Regions I and II, and between Regions II and III. Each of the regions may be expressed by the following equations:

Region I

$\begin{array}{cc}Iph<\frac{C_{\mathrm{FTP}}}{\mathrm{tac}1}\left(V_{\mathrm{FTP}\_M}-V_{\mathrm{FTP}\_L}\right)& \left(15\right)\\ \mathrm{Sig}=G_{\mathrm{SF}}\times \frac{I_{ph}}{C_{\mathrm{pix}}}\times \left(\mathrm{tac}1+\mathrm{tac}2\right)& \left(16\right)\end{array}$

Region II

$\begin{array}{cc}\frac{C_{\mathrm{FTP}}}{\mathrm{tac}1}\left(V_{\mathrm{FTP}\_M}-V_{\mathrm{FTP}\_L}\right)<\mathrm{Ip}h<\frac{C_{\mathrm{FTP}}}{\mathrm{tac}2}\left(V_{\mathrm{FTP}\_H}-V_{\mathrm{FTP}\_M}\right)& \left(17\right)\\ \mathrm{Sig}=\frac{G_{\mathrm{SF}}}{C_{\mathrm{pix}}}\times \left[C_{\mathrm{FTP}}\times \left(V_{\mathrm{FTP}\_M}-V_{\mathrm{FTP}\_L}\right)+I_{ph}\times \mathrm{tac}2\right]& \left(18\right)\end{array}$

Region III

$\begin{array}{cc}\frac{C_{\mathrm{FTP}}}{\mathrm{tac}2}\left(V_{\mathrm{FTP}\_H}-V_{\mathrm{FTP}\_M}\right)<\mathrm{Ip}h& \left(19\right)\\ \mathrm{Sig}=\frac{G_{\mathrm{SF}}}{C_{\mathrm{pix}}}\times C_{\mathrm{FTP}}\times \left(V_{\mathrm{FTP}\_H}-V_{\mathrm{FTP}\_L}\right)=\mathrm{saturation}& \left(20\right)\end{array}$

By controlling the integration periods tac1 and tac2, the sensitivity of each region can be controlled. Additionally, the output range in each region is controlled by the levels of FTP pulses. If tac1 should become shorter in relation to tac2, the sensitivity in Region II would become lower than that of Region I. Accordingly, the dynamic range of I_ph would increase, while the overall output dynamic range remained the same. Since the photoconversion of each region is linear, the image processing required for colored images becomes simplified. Also, the controlled photo response is independent of temperature, so that a more stable performance characteristic can be achieved, and that greater uniformity between pixel outputs can be achieved.

FIGS. 8 and 9 are an exemplary timing diagram and a photo conversion graph which illustrate how the circuit of FIG. 1A can be operated to achieve multiple knee points (900-902) through a simple modification in the FTP pulse. It should be understood that the number of “knee,” or transition points may be increased further by increasing the number of overflow pulses, and is not limited to the three-transition embodiment discussed herein. The timing diagram in FIG. 8 illustrates three integration periods: tac1 (t₆-t₅), tac2 (t₈-t₇) and tac3 (t₁′-t₉), where two different medium-level voltages (V_FTP—M1 and V_FTP—M2) are applied to the FTP pulses during the first (t₇-t₆) and second (t₉-t₈) overflow time periods. Under the exemplary embodiment of FIG. 8, the operation of the circuit is such that tac1>tac2>tac3.

The transition points of each “knee” are dependent upon the photodiode current I_ph that is produced after an integration period. Thus, each I_ph transition point (900-902) can be expressed as:

$\begin{array}{cc}{I}_{ph}\left(\mathrm{transition}1\right)=\frac{C_{\mathrm{FTP}}}{\mathrm{tac}1}\left(V_{\mathrm{FTP}\_M1}-V_{\mathrm{FTP}\_L}\right)& \left(21\right)\\ I_{ph}\left(\mathrm{transition}2\right)=\frac{C_{\mathrm{FTP}}}{\mathrm{tac}2}\left(V_{\mathrm{FTP}\_M2}-V_{\mathrm{FTP}\_M1}\right)& \left(22\right)\\ I_{ph}\left(\mathrm{transition}3\right)=\frac{C_{\mathrm{FTP}}}{\mathrm{tac}3}\left(V_{\mathrm{FTP}\_H}-V_{\mathrm{FTP}\_M2}\right)& \left(23\right)\end{array}$
Under a low illumination condition, if I_ph does not reach the level expressed in equation (21), no overflow current will subsequently flow in the 1st and 2nd overflow periods. In such a case, the pixel response may be expressed as:

$\begin{array}{cc}{V}_{\mathrm{PIX}}\left(t{1}^{\prime }\right)=V_{\mathrm{PIX}}\left(t_4\right)-\frac{I_{ph}\times \left(\mathrm{tac}1+\mathrm{tac}2+\mathrm{tac}3\right)}{C_{\mathrm{pix}}}& \left(24\right)\end{array}$

However, when I_ph exceeds the transition expressed in equation (21) after the first integration period, overflow current begins to flow during the first overflow period (t₇-t₆), and the resulting VPIX voltage is pinned by the overflow operation at t₇:

$\begin{array}{cc}{V}_{\mathrm{PIX}}\left(t_7\right)=V_{\mathrm{PIX}}\left(t_4\right)-\frac{C_{\mathrm{FTP}}}{C_{\mathrm{pix}}}\left(V_{\mathrm{FTP}\_M1}-V_{\mathrm{FTP}\_L}\right)& \left(23\right)\end{array}$

After the first overflow period, the pixel continues accumulation of charge during the second (tac2) and third (tac3) integration periods. The resulting V_PIX signal being read out at time t₁′ can be expressed as:

$\begin{array}{cc}{V}_{\mathrm{PIX}}\left(t_1^{\prime }\right)=V_{\mathrm{pix}}\left(t_4\right)-\frac{C_{\mathrm{FTP}}}{C_{\mathrm{pix}}}\left(V_{\mathrm{FTP}\_M1}-V_{\mathrm{FTP}\_L}\right)-I_{ph}\times \frac{\left(\mathrm{tac}2+\mathrm{tac}3\right)}{C_{\mathrm{pix}}}& \left(24\right)\end{array}$

As I_ph increases further and exceeds the second transition point, overflow current will flow in the second overflow period, pinning the V_PIX voltage at time t₉:

$\begin{array}{cc}{V}_{\mathrm{PIX}}\left(t_9\right)=V_{\mathrm{PIX}}\left(t_4\right)-\frac{C_{\mathrm{FTP}}}{C_{\mathrm{pix}}}\left(V_{\mathrm{FTP}\_M2}-V_{\mathrm{FTP}\_M1}\right)& \left(25\right)\end{array}$

thus:

$\begin{array}{cc}{V}_{\mathrm{PIX}}\left(t_1^{\prime }\right)=V_{\mathrm{PIX}}\left(t_4\right)-\frac{C_{\mathrm{FTP}}}{C_{\mathrm{pix}}}\left(V_{\mathrm{FTP}\_M2}-V_{\mathrm{FTP}\_L}\right)-I_{\mathrm{ph}}\times \frac{\mathrm{tac}3}{C_{\mathrm{pix}}}& \left(26\right)\end{array}$
If I_ph becomes sufficiently large and V_PIX reaches an overflow level after the third integration period, the photo conversion operation becomes saturated.

Turning to FIG. 9, the illustrated graph shows an exemplary photo conversion response of pixel circuit 100 operating under the timing shown in FIG. 8. The graph in FIG. 9 shows three different photo conversion gain responses (903-905), where each transition “knee” (900-902) results in the formation of Regions I-IV. Each of the regions may be expressed by the following equations:

Region I

$\begin{array}{cc}{I}_{\mathrm{ph}}<I_{\mathrm{ph}}\left(\mathrm{transition}1\right)& \left(27\right)\\ \mathrm{Sig}=G_{\mathrm{SF}}\times \frac{I_{\mathrm{ph}}}{C_{\mathrm{pix}}}\times \left(\mathrm{tac}1+\mathrm{tac}2+\mathrm{tac}3\right)& \left(28\right)\end{array}$

Region II

$\begin{array}{cc}{I}_{\mathrm{ph}}\left(\mathrm{transition}1\right)<I_{\mathrm{ph}}<I_{\mathrm{ph}}\left(\mathrm{transition}2\right)& \left(29\right)\\ \mathrm{Sig}=\frac{G_{\mathrm{SF}}}{C_{\mathrm{pix}}}\times \left[\begin{array}{c}{C}_{\mathrm{FTP}}\times \left(V_{\mathrm{FTP}\_M1}-V_{\mathrm{FTP}\_L}\right)+\\ I_{\mathrm{ph}}\times \left(\mathrm{tac}2+\mathrm{tac}3\right)\end{array}\right]& \left(30\right)\end{array}$

Region III

$\begin{array}{cc}{I}_{\mathrm{ph}}\left(\mathrm{transition}2\right)<I_{\mathrm{ph}}<I_{\mathrm{ph}}\left(\mathrm{transition}3\right)& \left(31\right)\\ \mathrm{Sig}=\frac{G_{\mathrm{SF}}}{C_{\mathrm{pix}}}\times C_{\mathrm{FTP}}\times \left(V_{\mathrm{FTP}\_M2}-V_{\mathrm{FTP}\_L}\right)+I_{\mathrm{ph}}\times \mathrm{tac}3& \left(32\right)\end{array}$

Region IV

$\begin{array}{cc}{I}_{\mathrm{ph}}>I_{\mathrm{ph}}\left(\mathrm{saturation}\right)& \left(33\right)\\ \mathrm{Sig}=\frac{G_{\mathrm{SF}}}{C_{\mathrm{pix}}}\times C_{\mathrm{FTP}}\times \left(V_{\mathrm{FTP}\_H}-V_{\mathrm{FTP}\_L}\right)=\mathrm{saturation}& \left(34\right)\end{array}$

Each of the photo response conversion gains 903-905 shown in FIG. 9 are determined by the integration periods tac1-tac3, where

$\mathrm{Gain}1=\frac{1}{\mathrm{tac}1+\mathrm{tac}2+\mathrm{tac}3},\mathrm{Gain}2=\frac{1}{\mathrm{tac}2+\mathrm{tac}3},\mathrm{and}\mathrm{Gain}3=\frac{1}{\mathrm{tac}3}.$
The ranges of the photo responses (906-908) between transition points (900-902) are a function of the FTP voltage, and may be expressed as Range1=(V_FTP—M1−V_FTP—L); Range2=(V_FTP—M2−V_FTP—M1) and Range3=(V_FTP—H−V_FTP—M2). Thus it can be seen that the range and gain of each region can be controlled by a predetermined pulse height and overflow timing of the FTP pulse, thus providing flexibility in optimizing photo conversion characteristics through the application of different FTP pulses.

Turning to FIG. 10, the block diagram illustrates an exemplary embodiment of an imager 1010 using the knee response pixel described above. The imager 1010 consists of a pixel array 1000, having n×m pixels, having a timing control block 1008, which provides driving and control pulses, along with sync signals to external circuits. The row address block 1006 generates row address pulses from address signals received from the timing controller 1008, and transmits the pulses to the level mix block 1007. Level Mix block 1007 generates row pulses, including RD, RST and FTP, for each of the rows (not shown) in the pixel array 1000.

The analog process block 1001 comprises of an amplifier array, a correlated double sampling (CDS) array and an analog memory array (which have been omitted for purposes of simplicity), where pixel outputs from pixel array 1000 are brought up to a required gain level, and where fixed pattern noise caused by variations in the pixel offset are suppressed by the CDS operation and stored in the analog memory array. The column address block 1002 receives column address signals from timing control block 1008, and generates column address pulses that are transmitted to the analog process block 1001, so that stored signals in the analog memory array may be read out. The signal readout from the analog memory is transmitted to the analog-to-digital converter (ADC) block 1003, where the signal is digitally converted and transmitted to the digital process block 1004 for processing (e.g., white balance, color interpolation, gamma correction, etc.). Once processed, the signal is then outputted from the output block 1005.

FIG. 11 illustrates an exemplary timing diagram for one frame cycle having a single “knee” point of the imager of FIG. 10. Signals RD, RST and FTP are output by the level mix block 1007 and are illustrated for each row line (1-m). The SHS and SHR pulses for CDR operation have been omitted for the purpose of clarity. The frame cycle for each row line in the embodiment of FIG. 11 begins when a respective row's RD pulses goes high, and ending when the RD pulse goes high again after two integration periods (tac1, tac2).

Turning to row line 1 of FIG. 11, the row address block 1006 outputs a rows select RD(1) pulse at the beginning of the frame cycle, and subsequently outputs a reset pulse RST(1) to reset the pixels in the first row. After the FTP(1) pulse is outputted, the first integration period tac1 is initiated, and continues until level mix block 1007 generates a short RD(1) pulse, ending the first integration period (tac1), and producing an overflow pulse FTP(1) during the horizontal blanking period of row line m-2.

A row cycle is illustrated in the exemplary embodiment of FIG. 11 between the rising edges of each row select pulses (RD(m-2), RD(m-1)) of adjacent row lines (row line m-2 and row line m-1), where each row cycle is comprised of a horizontal blanking period, followed by a data scanning period. Data stored in the analog memory array in the analog processing circuit 1001 are scanned and read out during the data scanning period, so that, for example, the overflow operation initiated for the first row does not affect data readout for the m-2 row.

Once a row's frame cycle is complete (e.g., row line 1), the operation moves sequentially to the next row (row line 2) to begin a new cycle, until all rows are read, reset and submitted to an overflow operation. Once the last row (row line m) is reached, one frame period will have been completed. In the exemplary embodiment of FIG. 11, signal integration periods tac1 and tac2 are held constant through each row, so that the same photo conversion characteristics and knee responses can be obtained in the entire pixel array region. It should be understood that, while a single “knee” point was described in the embodiment, that multiple knee points can be obtained by providing additional overflow pulses described in the embodiments above.

FIG. 12 illustrates an exemplary processing system 2000 which utilizes a pixel circuit such as that described in connection with FIGS. 1-11. The processing system 2000 includes one or more processors 2001 coupled to a local bus 2004. A memory controller 2002 and a primary bus bridge 2003 are also coupled the local bus 2004. The processing system 2000 may include multiple memory controllers 2002 and/or multiple primary bus bridges 2003. The memory controller 2002 and the primary bus bridge 2003 may be integrated as a single device 2006.

The memory controller 2002 is also coupled to one or more memory buses 2007. Each memory bus accepts memory components 2008. The memory components2008 may be a memory card or a memory module. The memory components 2008 may include one or more additional devices 2009. For example, in a SIMM or DIMM, the additional device 2009 might be a configuration memory, such as a serial presence detect(SPD) memory. The memory controller 2002 may also be coupled to a cache memory 2005. The cache memory 2005 may be the only cache memory in the processing system. Alternatively, other devices, for example, processors 2001 may also include cache memories, which may form a cache hierarchy with cache memory 2005. If the processing system 2000 includes peripherals or controllers which are bus masters or which support direct memory access (DMA), the memory controller 2002 may implement a cache coherency protocol. If the memory controller 2002 is coupled to a plurality of memory buses 2007, each memory bus 2007 may be operated in parallel, or different address ranges may be mapped to different memory buses 2007.

The primary bus bridge 2003 is coupled to at least one peripheral bus 2010. Various devices, such as peripherals or additional bus bridges may be coupled to the peripheral bus 2010. These devices may include a storage controller 2011, a miscellaneous I/O device 2014, a secondary bus bridge 2015, a multimedia processor 2018, and a legacy device interface 2020. The primary bus bridge 2003 may also be coupled to one or more special purpose high speed ports 2022. In a personal computer, for example, the special purpose port might be the Accelerated Graphics Port (AGP), used to couple a high performance video card to the processing system 2000.

The storage controller 2011 couples one or more storage devices 2013, via a storage bus 2020, to the peripheral bus 2010. For example, the storage controller 2011 may be a SCSI controller and storage devices 2013 may be SCSI discs. The I/O device 2014 may be any sort of peripheral. For example, the I/O device 2014 may be a local area network interface, such as an Ethernet card. The secondary bus bridge may be used to interface additional devices via another bus to the processing system. For example, the secondary bus bridge may be a universal serial port (USB) controller used to couple USB devices 2017 to the processing system 2000. The multimedia processor 2018 may be a sound card, a video capture card, or any other type of media interface, which may also be coupled to one additional device such as speakers 2019. The legacy device interface 2020 is used to couple legacy devices 2021, for example, older styled keyboards and mice, to the processing system 2000.

The processing system 2000 illustrated in FIG. 8 is only an exemplary processing system with which the invention may be used. While FIG. 8 illustrates a processing architecture especially suitable for a general purpose computer, such as a personal computer or a workstation, it should be recognized that well known modifications can be made to configure the processing system 2000 to become more suitable for use in a variety of applications. For example, many electronic devices which require processing may be implemented using a simpler architecture which relies on a CPU 2001 coupled to memory components 2008 and/or memory devices 2009. The modifications may include, for example, elimination of unnecessary components, addition of specialized devices or circuits, and/or integration of a plurality of devices.

While the invention has been described in detail in connection with preferred embodiments known at the time, it should be readily understood that the invention is not limited to the disclosed embodiments. Rather, the invention can be modified to incorporate any number of variations, alterations, substitutions or equivalent arrangements not heretofore described, but which are commensurate with the spirit and scope of the invention. Accordingly, the invention is not limited by the foregoing description or drawings, but is only limited by the scope of the appended claims.

Claims

1. A pixel circuit, comprising:

a photoconversion device for accumulating charge during a first integration period and second integration period;

an integration node, connected to said photoconversion device;

a first transistor, having one terminal connected to said integration node and another terminal connected to a reset signal line;

a feed-through pulse capacitor, having one terminal coupled to a feed-through pulse signal line, and a second terminal coupled to said integration node, said feed-through pulse signal line providing an intermediate pulse between said first and second integration periods to generate an overflow current in said first transistor.

2. The pixel circuit of claim 1, wherein said photoconversion device is a photodiode.

3. The pixel circuit of claim 1, wherein the first transistor has a gate coupled to a supply voltage.

4. The pixel circuit of claim 1, wherein said reset signal line and said feed-through pulse line are operable to apply a reset voltage to said integration node.

5. The pixel circuit of claim 4, further comprising an output circuit to read out said reset voltage to an output line.

6. The pixel circuit of claim 1, wherein a pixel select line coupled to the integration node, is operable to begin a first charge integration period.

7. The pixel circuit of claim 6, wherein said pixel select line is operable to begin a second charge integration period following said first integration period.

8. An imaging device, comprising:

a semiconductor substrate;

an array of pixels formed on said substrate, each pixel having a pixel circuit comprising:

a photoconversion device for accumulating a charge at an integration node;

a transfer transistor, having one source/drain terminal connected to said integration node and another source/drain terminal connected to a reset line;

an output circuit, connected to said integration node; and

a feed-through pulse capacitor, having one terminal coupled to a feed-through pulse signal line, and a second terminal coupled to said integration node, said feed-through pulse signal line producing an intermediate pulse that interrupts a charge integration period to create a first charge integration period and a second charge integration period.

9. The imaging device of claim 8, wherein said transfer transistor produces overflow current between said first and second charge integration periods in response to said intermediate pulse.

10. The imaging device of claim 8, wherein said photoconversion device is a photodiode.

11. The imaging device of claim 8, wherein said transfer transistor has a gate coupled to a supply voltage.

12. The imaging device of claim 8, wherein said reset signal line and said feed-through pulse line are operable to apply a reset voltage to said integration node.

13. The imaging device of claim 12, wherein said output circuit reads out said reset voltage to an output line.

14. An imaging system, comprising:

a plurality of pixels arranged in rows and columns, each pixel being associated with a column line, each pixel having a pixel circuit comprising:

a photodiode for accumulating charge during a first integration period and second integration period, said device being connected to an integration node;

a first transistor, having one terminal connected to said integration node and another terminal connected to a reset line;

an output circuit, connected to said integration node, for reading out a signal at said integration node; and

a feed-through pulse capacitor, having one terminal coupled to a feed-through pulse signal line, and a second terminal coupled to said integration node;

wherein an excess charge overflows through said first transistor in response to an intermediate pulse generated by said feed-through pulse signal line.

15. The imaging system of claim 14, wherein said excess charge overflows when a photocurrent present at said integration node after said first integration period exceeds a predetermined threshold.

16. The imaging system of claim 14, wherein the first transistor has a gate coupled to a supply voltage.

17. The imaging system of claim 14, wherein said reset signal line and said feed-through pulse line are operable to apply a reset voltage to said integration node.

18. The imaging system of claim 17, wherein said output circuit further comprises a second transistor for reading out said reset voltage to an output line.